Input Characteristics and Bioavailability After Administration of Immediate and a New Extended-Release Formulation of Hydromorphone in Healthy Volunteers David R

Home , First pass effect

Anesthesiology 2002; 97:827–36 © 2002 American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc. Input Characteristics and Bioavailability after Administration of Immediate and a New Extended-release Formulation of Hydromorphone in Healthy Volunteers David R. Drover, M.D.,* Martin S. Angst, M.D.,* Marta Valle, Ph.D.,† Bhamini Ramaswamy, M.D.,‡ Sujata Naidu, M.S.,§ Donald R. Stanski, M.D.,ʈ Davide Verotta, Ph.D.#

Background: To compare the pharmacokinetics of intrave- IN anesthesia clinical practice, intravenous bolus and nous, oral immediate-release (IR), and oral extended-release infusion is a familiar method of delivery for most drugs. (OROS®) formulations of hydromorphone. Methods: In this randomized, six-session, crossover-design The clinician knows precisely how much intravenous Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 study, 12 subjects received hydromorphone 8-mg intravenous, drug has been given and, with clinical experience and 8-mg IR oral, and 8-, 16-, and 32-mg OROS® formulations or pharmacologic understanding, can predict the clinical placebo orally followed by plasma sampling for hydromor- response from the given intravenous dosing regimen. phone determination. Pharmacokinetic analysis was per- With oral administration, the amount of drug delivered formed using NONMEM. Using the disposition of hydromorphone from the intravenous administration, deconvolution was systemically and the duration over which it is absorbed used to estimate the input rate function (release rate from the is difficult to assess clinically relative to intravenous gut to the blood) for the IR and OROS® formulations. A linear administration of the same drug. With oral administra- spline was used to describe the drug input rate function. tion, the absorption process and underlying bioavailabil- Results: The deconvolution using linear splines described the ity issues are complex, as the rate of drug entry into the in vivo release characteristics of both the IR and OROS® formulations. The mean absolute bioavailability for the 8-mg OROS® blood stream changes with time. When orally adminis- than for the tered sustained-release products are administered, this (0.025 ؍ formulation was significantly larger (P 8-mg IR formulation: 0.24 (SD 0.059) versus 0.19 (SD 0.054), issue becomes even more complex. Understanding the respectively. The bioavailability was the same for the three versus ® release rate profile time of an orally administered, doses of the OROS formulation. Predicted degree of fluctuation sustained-release product is analogous to understanding of plasma concentrations would be expected to be 130% and 39% for the IR and OROS® 8-mg doses, respectively. the infusion rate of an intravenously administered drug. Conclusions: The OROS® formulation of hydromorphone pro- This information allows more logical use of orally admin- duced continued release of medication over 24 h, which should istered drugs with sustained-release properties in the allow for once-daily oral dosing. The extended release of hy- patient population. dromorphone will produce less fluctuation of plasma concentrations compared with IR formulations, which should provide Hydromorphone is a derivative of morphine that is for more constant pain control. The in vivo release of hydro- approximately five times more potent and primarily acts morphone from both IR and OROS® formulations were ade- by binding to ␮-opioid receptors. Because of rapid elim- quately described using a linear spline deconvolution ap- ination and redistribution of hydromorphone, oral dos- ® proach. The increased bioavailability from the OROS ing every 4 h with the conventional immediate-release formulation may be related to decreased metabolism by a first- pass effect or enterohepatic recycling of hydromorphone. (IR) tablet is required to sustain adequate plasma concentrations.1 Sustained-release drug formulations provide more convenient dosing, improve compliance, and Additional material related to this article can be found on the produce more stable plasma concentrations, which Anesthesiology Web site. Go to the following address, click on should result in a more desirable therapeutic effect.2–5 the Enhancements Index, and then scroll down to find the Optimal treatment of chronic continuous pain requiring appropriate article and link. http://www.anesthesiology.org opioid analgesics is better managed with sustained plasma concentration of the opioid analgesic. A new once-daily extended-release formulation of hydromor- * Assistant Professor, ‡ Research Fellow, § Research Assistant, ʈ Professor, De- partment of Anesthesia, Stanford School of Medicine. † Research Fellow, Depart- phone has been developed for the treatment of chronic ment of Biopharmaceutical Sciences, # Associate Professor, Departments of Bio- pain. Osmotic pump delivery technology for the ex- pharmaceutical Sciences, Laboratory Medicine, Epidemiology and Biostatistics, and Pharmaceutical Chemistry, University of California, San Francisco, California. tended release of orally administered drugs has become ® Received from the Department of Anesthesia, Stanford University School of available during the past two decades. The OROS sys- Medicine, Stanford, California. Submitted for publication February 22, 2001. tem (ALZA Corporation, Mountain View, CA) previously Accepted for publication June 3, 2002. Supported by a grant from Abbott 6 Laboratories, Abbott Park, Illinois. Presented in part at the annual meeting of the described by Theeuwes consists of a semipermeable American Society of Anesthesiologists, Dallas, Texas, October 9–13, 1999. Knoll membrane surrounding a bilayer tablet core, one layer Pharmaceutical Company, North Mount Olive, New Jersey, sponsored the orig- inal study and was sold to Abbott Laboratories during preparation of the manu- containing the drug and the other an osmotically active script. Dr. Drover has a consulting agreement with Abbott Laboratories that is not component. A small hole is drilled through the mem- related to this study or any of these formulations. brane on the side adjacent to the drug layer. In the Address correspondence to Dr. Drover: Department of Anesthesia, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, California 94305-5640. gastrointestinal tract, water diffuses across the mem- Address electronic mail to: [email protected]. Reprints will not be available from the authors. Individual article reprints may be purchased through brane, and a gel-like suspension is formed in the drug the Journal Web site, www.anesthesiology.org. layer. As the osmotic layer expands, the drug suspension

Anesthesiology, V 97, No 4, Oct 2002 827 828 DROVER ET AL. is pushed through the orifice at a near constant rate into given over 10 min. Arterial blood samples for pharma- the lumen of the gastrointestinal tract for absorption. cokinetic analysis were drawn at 0 (prior to dosing), 1, 3, The goals of this study were to determine the release 5, 7, 10, 11, 12, 13, 15, 17, 20, 25, 40, 55, 70, 100, 130, characteristics and bioavailability of an OROS® sus- 190, 250, and 370 min after commencing the infusion. A tained-release hydromorphone preparation. To accu- second intravenous catheter was placed for sampling of rately understand the release rate of a sustained-release later blood specimens at 370, 490, 730, 970, and 1,450 product, it is necessary to separate the drug absorption min after dosing. Simultaneous venous and arterial sam- phase from the drug distribution and elimination phases. ples were obtained at the 370-min sample time to con- To do this precisely, it is necessary to first measure the firm comparability of sampling sites. In all stages of the distribution and elimination using the intravenous route, study, the exact time of the midpoint of each sample then in the same individual, on another occasion, to draw was used as the time of the pharmacokinetic spec- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 administer the oral dosage form. Mathematical data anal- imen. Meals were taken at standardized times, i.e., 250 ysis can then separate the oral absorption phase of drug and 490 min after drug administration (coinciding blood entry into the blood stream from concurrent distribution drawings preceded the meals). and elimination phases. This approach allows compari- In the second stage, each subject received a single son of hydromorphone bioavailability from IR and dose of 8 mg IR hydromorphone (Dilaudid®; Abbott OROS® formulations and considers the effect of sus- Laboratories, Abbott Park, IL). Subjects swallowed the tained-release products on bioavailability. Understanding single tablet with a standardized amount of water the release characteristics of the sustained-release opiate (240 ml). No other water intake was allowed from 1 h will allow more logical use of this formulation in patients prior to dosing through 1 h after dosing. Meals were with chronic pain. taken at standardized times, i.e., 180, 540, and 1440 min after drug intake (coinciding blood drawings preceded the meals). Venous blood samples were drawn immedi- Materials and Methods ately before drug intake and 15, 30, 45, 60, 90, 120, 180, 240, 300, 360, 480, 720, 960, and 1,440 min after drug Subjects intake. After we obtained approval from the Stanford Univer- Dosing for each of the four periods in stage 3 (sessions sity Institutional Review Board, 12 healthy subjects 3–6) was randomized and double-blind; subjects re- (6 men and 6 women) were enrolled after giving written ceived single doses of 8, 16, and 32 mg OROS® hydro- informed consent. Subjects were included if they were morphone (Abbott Laboratories) and placebo in a ran- nonsmokers and consumed the caffeine equivalent of no domized sequence. A schematic representation of the more than three cups of coffee per day. Prior to enroll- OROS® release mechanism has been published.7 A dou- ment, all subjects had a screening physical examination, ble-blind, randomized, placebo-controlled design was 12-lead electrocardiogram, routine laboratory profile, used to minimize bias during acquisition of pharmaco- and drug screen. Women were required to test negative dynamic data.7 To maintain the blind, at the beginning of for pregnancy. The drug abuse screen and pregnancy each study session subjects swallowed three tablets, all test were confirmed negative prior to each study day. No containing no drug (placebo period) or, alternatively, alcohol or over-the-counter medication was allowed two containing no drug and one containing 8, 16, or within 48 h prior to each study day. Prescription drugs 32 mg of the OROS® formulation of hydromorphone. and chronic medications were prohibited during and Water intake was standardized as outlined for IR stage 2 14 days prior to the study except for oral contraceptives. sessions. Meals were taken at standardized times, i.e.,3, Prior to each study session, subjects fasted overnight. 9, 24, 28, 33, and 48 h after drug intake (coinciding blood drawings preceded the meals). Venous blood sam- Study Design ples were drawn before drug intake and 1, 2, 3, 6, 9, 12, A three-stage (six-session), placebo-controlled, cross- 15, 18, 21, 24, 30, 36, and 48 h after drug intake. over study design was used. The pharmacodynamic component of this study has been previously published.7 For Hydromorphone Assay the first stage, after subjects arrived at the study center, Seven-milliliter samples of blood were drawn into hep- electrocardiogram and hemoglobin oxygen saturation arinized glass tubes and centrifuged, and plasma was were continuously monitored, and a catheter was in- frozen within 1 h. Plasma was stored in polypropylene at serted into a radial artery for continuous hemodynamic Ϫ20°C. Plasma (1.0 ml) and the internal standard (tri- monitoring and to obtain blood samples for assessment deuterated hydromorphone) were extracted into or- of plasma drug concentration. An intravenous catheter ganic solvent. Following centrifugation, an aliquot of the was placed in the contralateral arm for administration of organic layer was injected onto a SCIEX APIIII-Plus LC- drug. After obtaining a baseline blood sample, an 8-mg MS-MS (PE Biosystems, Foster City, CA) using a short zero-order infusion (0.8 mg/min) of hydromorphone was high-performance liquid chromatography column. The

Anesthesiology, V 97, No 4, Oct 2002 PHARMACOKINETICS OF IV, IR, AND OROS HYDROMORPHONE 829 peak area of the m/z 2863185 product ion of hydro- Equation 1 represents a causal linear time-invariant morphone was measured against the peak area of the system, one that depends only on input up to and in- m/z 2893185 product ion of the internal standard. cluding time (t). The response of the system at time Quantitation was performed using an xϪ1-weighted lin- t [R(t)] is described by the convolution of the disposition ear least-square regression line generated from spiked function [K(t)] with the input rate function [I(t)], which plasma calibration standards. The assay was linear over a is represented by a spline function. In this equation, t is range of 0.05 to 10.0 ng/ml. The between-day coefficient time and ␶ is an integration variable, the parameter ⑀ of variation ranged from 1.4 to 11.2% (10 runs). The limit denotes error between the model and the data. The two of quantification was 0.05 ng/ml. constraints used in this analysis were that absorption of hydromorphone would be nonnegative and that during the period after peak absorption, the curve would be Data Analysis Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 monotonically decreasing. Constraints are a method of Plasma concentration and time data were analyzed by providing information to the solution of the problem nonlinear regression using NONMEM version V (NONMEM without making unnecessary assumptions. The term Project Group, University of California, San Francisco, “spline” indicates a piece-wise polynomial function. A CA).8 As the first step, the disposition function of hydro- spline connects points corresponding to certain loca- morphone after intravenous administration was deter- tions called breakpoints or nodes. There are several mined for each subject. The frequency of data sampling options for choice of breakpoints: the breakpoints were allowed individual fits of the data for the purpose of chosen at the quantiles of the observation times.12 The obtaining the individual disposition parameters. One-, flexibility of a spline increases with the number of break- two-, and three-compartment mamillary models were points; thus, few breakpoints might fail to recover the considered for the disposition of hydromorphone. A correct input function, but too many might show arti- proportional error model was used, of the form y ϭ f facts (e.g., generate unrealistic oscillations in the spline). (1 ϩ eps), where y is an observation, f is the correspond- The number of breakpoints was determined both by ing prediction, and eps represents the error, which is visual assessment of the fit as well as using the Hannan assumed to be normally distributed with mean zero and criterion.9 unknown variance (the NONMEM software estimates The OROS® release function was also characterized in the unknown parameters in the function f and the error the same fashion. The breakpoints were chosen at the variance using maximum likelihood). Best fit of the data quantiles of the observation times. The model selection for each individual subject was judged based on visual was based on the visual inspection of the residual error assessment of the data and the value of the Hannan plots and the value of the Hannan criteria.9 For a com- criteria.9 The estimated disposition parameters for each parison analysis, the IR formulation was considered the individual were used in the next steps of deconvolution. reference and the OROS® formulation the test com- Deconvolution describes the mathematical separation of pound. Bioavailability was calculated using the area un- the absorption phase from the distribution and elimina- der the input rate function curve as determined by the tion phases. linear spline analysis performed with NONMEM. For Hydromorphone IR deconvolution was initially per- validation of the area under the input rate function formed using a traditional parametric approach with curve, the area under the curve (AUC) was also calcu- NONMEM, using the disposition function from the intra- lated from observed concentrations versus time by the venous hydromorphone period of the study and assum- log-linear trapezoidal rule. Statistical comparison of bio- ing first-order absorption. A lag time for absorption was availability data was performed using analysis of variance also considered as a component of this analysis. The on the logarithmically transformed data.13 deconvolution was subsequently repeated using a semi- To gain a clinical impression of the mean behavior of parametric approach. Instead of assuming first-order ab- the concentration profile of hydromorphone after dos- sorption, the absorption phase was described using a ing with 8-mg IR and OROS® formulations, a simulation linear spline10–12 to characterize the input rate function. was performed with repeating doses of each formula- A spline is a mathematical function that describes a tion, to a maximum of eight doses, which assures attain- smooth curve connecting a set of points (e.g., break- ment of steady state. This calculation is only an estima- points, nodes) and was chosen because a spline de- tion and is based on the following assumptions: (1) the scribes an arbitrary shape. This is described here by the pharmacokinetics in a patient population are similar to following equation: this volunteer population; (2) hydromorphone terminal

t phase of the elimination curve is accurately estimated; and (3) multiple doses do not change the pharmacoki- R(t)ϭ͵ K(␶) ⅐ I(␶ Ϫ t)dt ϩ ␧ (1) netics of hydromorphone. Simulation was performed using NONMEM for both 8-mg IR and 8-mg OROS® 0 hydromorphone formulations. A simulation considering

Anesthesiology, V 97, No 4, Oct 2002 830 DROVER ET AL.

Table 1. Bioavailability Data

Bioavailability

Subject IR OROSா

1 0.15 0.186 2 0.246 0.381 3 0.197 0.252 4 0.131 0.214 5 0.264 0.313 6 0.214 0.203 7 0.153 0.263 8 0.214 0.248

9 0.157 0.231 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 10 0.106 0.184 11 0.274 0.241 12 0.161 0.177

Mean 0.189 0.241 SD 0.054 0.059

Bioavailability was calculated using the area under the input rate function curve. IR ϭ immediate release; OROSா ϭ oral extended-release formulation.

interindividual variability was not performed because of limited information obtained from the small study pop- Fig. 1. Measured hydromorphone concentrations from each of ulation (12 subjects). IR was simulated based on every the 12 subjects versus time after intravenous infusion of 8 mg 8-h dosing intervals and OROS® based on every 24-h hydromorphone given over 10 min. The insert expands the dosing interval, which may be reasonable clinical use first1h. projections for these two formulations. A total daily dose of 48 mg was used such that each formulation would be best described using a three-compartment model Ͻ simulated based on the same total daily dose. At steady (P 0.001, in respect to a two-compartment model) state, the degree of plasma concentration fluctuation with first-order elimination from the central compart- (DF) was estimated for both formulations as follows: ment. Raw plasma concentration-versus-time data for intravenous hydromorphone is presented graphically in Ϫ (Cmax Cmin)*100 figure 1. The individual parameters of the disposition DF ϭ (2) Cavg function are available (table 2 web enhancement; additional information is available on the ANESTHESIOLOGY Web where C is the maximum predicted plasma concentra- max site at http://www.anesthesiology.org). The unit dispo- tion, C is the minimum predicted plasma concentration, min sition functions obtained from the three-compartment and C is the average predicted plasma concentration at avg model for the 12 subjects are available (fig. 1 web en- steady state. hancement; additional information is available on the ANESTHESIOLOGY Web site at http://www.anesthesiology- Results .org). These data were then used as the individual disposition function for further deconvolution of data from Twelve subjects (6 male, 6 female), aged 21–34 yr, the oral dosage forms using NONMEM. were enrolled in the study. All subjects completed all study sessions (table 1 web enhancement; additional Stage 2: Immediate-release Oral Hydromorphone information is available on the ANESTHESIOLOGY Web site at The disposition function from the intravenous data http://www.anesthesiology.org). The side effects, as re- were fixed for each individual to estimate the IR input ported previously,7 were consistent with those associ- rate function. The raw concentration–time data used in ated with opioid analgesics, the most common of which the calculations are displayed graphically in figure 2. were nausea, pruritus, and lightheadedness. After the Although initial analysis considered a model with first- intravenous administration of 8 mg hydromorphone, no order absorption and elimination, visual inspection of subjects had loss of consciousness, apnea, or a hemoglo- the residual error plots and individual plasma concentra- bin oxygen saturation less than 90%. tions indicated model misspecification. Residual error plots showing measured over predicted concentrations Stage 1: Intravenous Hydromorphone versus time were constructed for the spline input and Each individual subject’s plasma concentration-versus- first-order absorption models (fig. 3) to demonstrate the time data were fit separately using NONMEM and were degree of model misspecification. Deconvolution using a

Anesthesiology, V 97, No 4, Oct 2002 PHARMACOKINETICS OF IV, IR, AND OROS HYDROMORPHONE 831 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021

Fig. 2. Measured hydromorphone concentrations from each of 12 subjects versus time after oral administration of 8 mg immediate-release (IR) hydromorphone. The insert expands the first5h. linear spline was then performed to estimate the input Fig. 3. Measured/predicted plasma drug concentrations from rate function. The corresponding fit resulted in a good each of 12 subjects after oral administration of 8 mg immediate- release (IR) hydromorphone. The y-axis displays the ratio of description of the data and revealed that hydromor- measured to predicted drug concentrations for all subjects. The phone from the IR formulation is not immediately re- predicted concentrations were calculated from the pharmaco- ؍ leased and absorbed after dose administration. Figure kinetic model as estimated by NONMEM. The line drawn at y 1 represents perfect prediction. (Top) The model using a linear 4 shows the input rate functions for each of the 12 spline input function; (bottom) data of the model assuming subjects. There was an initial rapid release followed by first-order absorption. second phase of slower hydromorphone release from the IR tablet. The mean value of the maximum rate of (SD 0.38). The bioavailability per milligram oral dose was release was 17 mg/h (SD 10). Absolute hydromorphone the same for 8-, 16-, and 32-mg OROS® hydromorphone bioavailability values (F) calculated from the area under doses, indicating proportional drug release and absorp- the spline input rate-versus-time curve for each subject tion profiles with the three tablet strengths. To better (mean F ϭ 0.189 [SD 0.054]) are presented in table 1. display the time of peak plasma concentration and vari- The analogous bioavailability determined from AUC us- ability of the three OROS® doses, the 90% confidence ing the trapezoidal rule (mean F ϭ 0.194 [SD 0.050]; data limit of the parameters was calculated, and representa- not shown) was 2.8% greater (not significant, P Ͼ 0.05) tive curves were produced. Figure 7 shows pointwise than that determined using the AUC under the spline 90% confidence intervals for the OROS input and corre- input rate curve. sponding predictions. The pointwise 90% confidence intervals were obtained using the mean disposition func- Stage 3: Oral Hydromorphone OROS® Formulation tion convolved with the 5% and 95% OROS input pro- The raw data from all subjects for each of the three files; thus, they show the effect of variability in the input doses are displayed in figure 5. Figure 6 shows the input rate on drug concentrations. The bioavailability of hy- rate functions obtained using deconvolution of the dromorphone from the OROS® formulation calculated OROS® hydromorphone formulation data for each sub- from the area under the input rate-versus-time curve ject. The input function suggests continued release of (mean F ϭ 0.241 [SD 0.059]) for each individual is hydromorphone from the OROS® delivery system over presented in table 1. Bioavailability for each individual at the sampling period, with the maximum absorption rate each dose is available (table 3 web enhancement; addi- detected at approximately 10 h and decreasing progres- tional information is available on the ANESTHESIOLOGY Web sively thereafter. The mean value of the maximum rate of site at http://www.anesthesiology.org). The hydromor- release for the 8-mg OROS® formulations was 1.2 mg/h phone bioavailability from the OROS® formulation cal-

Anesthesiology, V 97, No 4, Oct 2002 832 DROVER ET AL. Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021

Fig. 4. Estimated individual input rate functions for 8 mg immediate-release (IR) hydromorphone for each of 12 subjects. Subject identiﬁcation numbers are included in each plot.

culated using the input rate function was significantly degree of fluctuation would be expected to be 130% for greater than that of the IR formulation (P ϭ 0.025). The the IR and 39% for the 8-mg OROS® dose. analogous bioavailability determined from AUC using the trapezoidal rule (mean F ϭ 0.251 [SD 0.073]) was 4.3% greater (not significant) than the AUC under the spline Discussion input rate curve. Physicians use opioids to treat patients with pain con- Simulation of Typical Dosing of Immediate-release ditions. In patients with moderate to severe chronic and OROS® Hydromorphone pain, ideally a steady state plasma concentration from an Multiple dosing was simulated using the mean values effective and well-tolerated dose of analgesic medication from the individual fits from the NONMEM analysis for is achieved. Historically, frequent dosing with short-act- both IR (fig. 8) and OROS® (fig. 9) hydromorphone. The ing analgesic formulations has lead to continued or epi- predicted plasma concentrations suggest that the peak sodic breakthrough pain, resulting in a potential de- plasma concentration would be almost 37% greater after crease in patient compliance and satisfaction.1,14 Orally an IR dose than an OROS® dose (9.2 vs. 6.7 ng/ml). The administered sustained-release opioid formulations have

Anesthesiology, V 97, No 4, Oct 2002 PHARMACOKINETICS OF IV, IR, AND OROS HYDROMORPHONE 833

with appearance of drug in the sampling compartment would not be a satisfactory mathematical model when drug release, absorption, and elimination occur simulta- neously and in a complex fashion. Drug input into the systemic circulation is a factor of both the release of drug from the dosage form and rate of absorption across the mucosal wall. Since hydromorphone from the OROS® formulation can be released along the entire length of the gastrointestinal tract, the absorption rate constant may not be consistent in all segments of the tract. When the IR formulation was modeled assuming first- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 order absorption characteristics, there was significant model misspecification. This would suggest that even drugs that are considered IR might have a delayed-release component because of the inherent composition of the formulation. The spline method of deconvolution also showed a degree of model misspecification (fig. 3) but was the superior model when compared with a first-order absorption model, with or without a lag time. Compressed tablet formulations have been shown to have more variable release characteristics than formulations using an “effervescent” system.15 Effervescent or liquid formulations might be most appropriately described as IR and show good mathematical fittoafirst- order absorption model. The intravenous form of a drug may not always be available or safe to administer to human subjects. While Fig. 5. Measured hydromorphone concentrations versus time it may be tempting to use “known” pharmacokinetic ® after oral administration of 8, 16, and 32 mg of OROS parameter estimates from an “IR” formulation in the hydromorphone. deconvolution of a sustained-release formulation, those been developed that can provide the patient with a more assumptions may not always be appropriate. In the cur- convenient dosing regimen and with less variable plasma rent study, when data from the IR formulation were used opioid concentrations than may be possible with re- to provide parameters (e.g., Ka) for the deconvolution of peated dosing of IR formulations. In this clinical investi- the OROS® formulation of hydromorphone, the errone- gation, we studied a once-daily oral OROS® hydromor- ous impression that the OROS® formulation had nonlin- phone formulation that produces a controlled drug ear kinetics was concluded (data not shown). The cur- release over the 24-h dosing interval. This hydromor- rent spline input curves describe a convolution of the phone preparation has been shown to provide sustained release characteristics of hydromorphone from the analgesia after a single daily dose in an experimental pain OROS® formulation and the resultant Ka. If an accurate model.7 The mathematical model used here effectively Ka could have been calculated from the IR deconvolu- described the release characteristics of hydromorphone tion, a further deconvolution calculation could have al- from this new sustained-release formulation. The OROS® lowed exact determination of the in vivo release profile formulation produces sustained plasma concentrations of hydromorphone from the OROS® formulation. How- over a 24-h interval that would be consistent with ever, the IR kinetics could not be adequately described expectations for effective once-daily dosing with by a first-order absorption constant, and it was therefore hydromorphone. not possible to completely characterize the in vivo re- Medications are being prepared in new formulations to lease profile from this extended-release formulation. make patient dosing easier and to improve drug effect, The absorption of hydromorphone for 24 h from the tolerance, and patient compliance. Delivering short-act- OROS® formulation may be partially explained by the ing drugs in extended-release systems can produce a presence of enterohepatic cycling of hydromorphone. long-lasting effect with infrequent dosing. However, As has been described for morphine, unchanged drug is when a short-acting drug such as hydromorphone is excreted in the bile and reabsorbed across the intestinal administered in a slow-release delivery system, the drug wall into the systemic circulation. This can give the may not appear in the circulation in a pattern that can be appearance of continued release from an oral formula- described effectively by the traditional sum-of-exponen- tion and produce a calculation of a greater bioavailability tials model. A first-order absorption process concurrent than an IR formulation.16–18

Anesthesiology, V 97, No 4, Oct 2002 834 DROVER ET AL. Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021

Fig. 6. Estimated individual input rate functions for 8 (lowest line), 16 (middle line), and 32 mg (upper line) OROS® formulation of hydromorphone for each subject. Subject identiﬁcation numbers are included in each plot.

The bioavailability of hydromorphone from other oral liver, where they may undergo extensive metabolism formulations in humans has been reported previous- before entering the rest of the system; this has been ly.19–21 The greater bioavailability reported previously termed the “first-pass effect.” The OROS® hydromor- cannot be explained by this study. Previous studies in phone formulation had a greater bioavailability than the humans have used shorter sampling periods than our IR form. This might be explained by the release and study. This sampling difference may account for a differ- absorption of hydromorphone from the OROS® formu- ent pharmacokinetic model and might lead to incorrect lation in the colon and rectum, thus avoiding or decreas- estimation of the AUC. Also, previous studies used a ing the amount of the first-pass effect. Similarly, the radioimmunoassay method of quantitation of hydromor- OROS® formulation of oxybutynin has been found to phone that was less specific for the parent drug and may have a greater bioavailability than the IR oxybutynin have detected glucuronide metabolites. The absolute formulation, and a difference in first-pass metabolism bioavailability of hydromorphone implies a significant was postulated as the likely cause for this greater bio- amount of first-pass metabolism by the liver. Rapidly availability.2 Conversely, the extended-release formula- released drugs that are absorbed from the duodenum tions may contain undelivered drug at the time of dosage enter the mesenteric vessels and are transported to the form excretion in the stool. This would be expected to

Anesthesiology, V 97, No 4, Oct 2002 PHARMACOKINETICS OF IV, IR, AND OROS HYDROMORPHONE 835 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021

Fig. 9. Simulation of repeated doses of hydromorphone OROS® (48 mg) given orally every 24 h.

what plasma concentrations would be produced when these two different dosage forms are used clinically. The Fig. 7. Simulated curves of the 8-, 16-, and 32-mg OROS® doses of IR formulation may be expected to produce a greater hydromorphone. The dotted lines above and below each mean peak concentration of an equivalent dose from the dose curve represents the area of the 90% confidence interval as ® calculated from the parameter estimates. OROS formulation. The degree of fluctuation of plasma hydromorphone is projected to be far greater with the IR decrease relative bioavailability. In addition, the long (130%) than the OROS® (39%) formulation, even with period of release noted for this formulation (24–48 h) the more frequent dosing of the IR formulation. Since might be considered longer than most patient’s normal the degree of fluctuation is considerably larger with the gastrointestinal tract transit time. IR formulation, clinically it may translate to more side Two possible reasons for this apparently long release effects from greater peak plasma concentrations or a period exist. First, hydromorphone, like most opioids, greater likelihood for breakthrough pain between sched- delays gastric motility and produces constipation; thus, uled doses of medication. From the simulation, although ® OROS may reside in the gastrointestinal tract longer the extended-release formulation showed very slow re- than other medications. Second, decreased sampling fre- lease and delayed time to peak plasma concentration, by quency during the 24–48 h postdose period would pro- the third dose (third day) the peak plasma concentration duce model error in predicting final release characteris- would be expected to be 95% of the steady state peak. tics from the formulation. This error results in an This is useful clinical information for adjusting dosage inability to accurately determine whether plasma con- with the OROS® formulation. Dose adjustments with centrations at the later sampling points are a result of this extended-release formulation at 3-day intervals ap- continued release or final disposition of the drug; more pear reasonable. Steady state plasma concentrations are intense sampling during this interval may have improved likely attained after 3 days, providing the clinician with prediction of plasma concentrations. adequate information to evaluate and balance efficacy ® The characteristics of the IR and OROS formulations with tolerability before considering further titration. were simulated mathematically to gain an impression of In summary, hydromorphone when given intravenously was best described by a three-compartment pharmacokinetic model. The release of drug from IR hydromorphone was better described by a spline function than a first-order absorption model. The OROS® hydromorphone formulation showed an extended-release profile with much less peak-to-trough variation than the IR formulation. The bioavailability of hydromorphone from the OROS® formulation appeared greater than from the IR formulation, which may be a result of a difference in first-pass effect. Overall, the extended-release profile of hydromorphone from the OROS® formulation appears suitable for true once-daily oral dosing.

The authors thank Lawrence Saidman, M.D. (Professor, Department of Anes- thesia, Stanford School of Medicine, Stanford, California), Tom Valente, M.D. Fig. 8. Simulation of repeated doses of hydromorphone imme- (Abbott Laboratories, Abbott Park, Illinois), and Ralph Doyle, B.A. (Abbott Lab- diate-release (IR; 16 mg) given orally every 8 h. oratories, Mt. Olive, New Jersey) for reviewing the manuscript.

Anesthesiology, V 97, No 4, Oct 2002 836 DROVER ET AL.

References describe distribution, identify input, and control endogenous substances and drugs in biological systems. Crit Rev Biomed Eng 1996; 24:73–139 1. Bruera E, Sloan P, Mount B, Scott J, Suarez-Almazor M: A randomized, 13. Steinijans VW, Diletti E: Statistical analysis of bioavailability studies: Para- double-blind, double-dummy, crossover trial comparing the safety and efficacy of metric and nonparametric confidence intervals. Eur J Clin Pharmacol 1983; oral sustained-release hydromorphone with immediate-release hydromorphone 24:127–36 in patients with cancer pain. Canadian Palliative Care Clinical Trials Group. J Clin 14. Hays H, Hagen N, Thirlwell M, Dhaliwal H, Babul N, Harsanyi Z, Darke AC: Oncol 1996; 14:1713–7 Comparative clinical efficacy and safety of immediate release and controlled 2. Gupta SK, Sathyan G: Pharmacokinetics of an oral once-a-day controlled- release hydromorphone for chronic severe cancer pain. Cancer 1994; release oxybutynin formulation compared with immediate-release oxybutynin. 74:1808–16 J Clin Pharmacol 1999; 39:289–96 15. Lotsch J, Kettenmann B, Renner B, Drover D, Brune K, Geisslinger G, 3. Gupta SK, Shah JC, Hwang SS: Pharmacokinetic and pharmacodynamic Kobal G: Population pharmacokinetics of fast release oral diclofenac in healthy characterization of OROS and immediate-release amitriptyline. Br J Clin Pharma- volunteers: Relation to pharmacodynamics in an experimental pain model. col 1999; 48:71–8 Pharm Res 2000; 17:77–84 4. Kendall MJ, Maxwell SR, Sandberg A, Westergren G: Controlled release 16. Westerling D, Frigren L, Hoglund P: Morphine pharmacokinetics and metoprolol: Clinical pharmacokinetic and therapeutic implications. Clin Pharma- effects on salivation and continuous reaction times in healthy volunteers. Ther cokinet 1991; 21:319–30 Drug Monit 1993; 15:364–74 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/97/4/827/407023/0000542-200210000-00013.pdf by guest on 28 September 2021 5. Greenberg RN: Overview of patient compliance with medication dosing: A 17. Poulain P, Hoskin PJ, Hanks GW, O AO, Walker VA, Johnston A, Turner P, literature review. Clin Ther 1984; 6:592–9 Aherne GW: Relative bioavailability of controlled release morphine tablets (MST 6. Theeuwes F: Elementary osmotic pump. J Pharm Sci 1975; 64:1987–91 continus) in cancer patients. Br J Anaesth 1988; 61:569–74 7. Angst MS, Drover DR, Lotsch J, Ramaswamy B, Naidu S, Wada DR, Stanski 18. Hasselstrom J, Sawe J: Morphine pharmacokinetics and metabolism in DR: Pharmacodynamics of orally administered sustained-release hydromorphone humans: Enterohepatic cycling and relative contribution of metabolites to active in humans. ANESTHESIOLOGY 2001; 94:63–73 opioid concentrations. Clin Pharmacokinet 1993; 24:344–54 8. NONMEM Project Group: NONMEM’s Users Guide. Edited by Beal S, Shei- 19. Vallner JJ, Stewart JT, Kotzan JA, Kirsten EB, Honigberg IL: Pharmacoki- ner L. San Francisco, University of California San Francisco, 1998 netics and bioavailability of hydromorphone following intravenous and oral 9. Hannan E: Rational transfer function approximation. Stat Sci 1987; administration to human subjects. J Clin Pharmacol 1981; 21:152–6 2:135–61 20. Parab PV, Ritschel WA, Coyle DE, Gregg RV, Denson DD: Pharmacokinet- 10. Park K, Verotta D, Gupta SK, Sheiner LB: Passive versus electrotransport- ics of hydromorphone after intravenous, peroral and rectal administration to facilitated transdermal absorption of ketorolac. Clin Pharmacol Ther 1998; 63: human subjects. Biopharm Drug Dispos 1988; 9:187–99 303–15 21. Ritschel WA, Parab PV, Denson DD, Coyle DE, Gregg RV: Absolute bio- 11. DeBoor C: A Practical Guide to Splines. New York, Springer-Verlag, 1978 availability of hydromorphone after peroral and rectal administration in humans: 12. Verotta D: Concepts, properties, and applications of linear systems to Saliva/plasma ratio and clinical effects. J Clin Pharmacol 1987; 27:647–53

Anesthesiology, V 97, No 4, Oct 2002